444 AIASKAN GLACIER STUDIES making severe inroads on the glacier, as it does upon the Childs and the southern; portion of Miles Glacier. No icebergs are discharged, for the glacier front is everywhere mantled by moraine and vegetation. In spite of the narrowing of the river from. 8400 feet just north of Allen Glacier to 400 feet in Baird Canyon the current is not quickened sufficiently to make the river undercut the glacier front, although the ice ia known to extend out to the river itself. Evidence of Flowage in Atten Glacier. The layers of ice in many portions of the bulb of Allen Glacier are not horizontal, or inclined upward at a low angle, as near the termini of many valley glaciers. They are complexly folded and crumpled and, at many points within the interior flat, the general dip is vertical. This supports the idea of uprising ice-currents within interior flats. The folds in several places (PI. CLXXVII) were found to< be thickened on crests and thinned on the limbs as in the type of what Van Hise has called similar folds.1 As rock folds of this type are taken as evidence of rock flowage it seems possible that these ice folds may be interpreted as evidence of glacier movement by actual viscous flowage. The faults which cut certain folds are later than the folds and seem to have been made in the zone of fracture, or crevassing, after ablation of the glacier surface removed the overlying ice so that the folded ice was in the zone of fracture rather than in the zone of flowage. The ice is largely made up of coarse crystals. History of Allen Glacier. The early stage of Allen Glacier as a tributary of the main Copper River Glacier seems to have been followed by a stage of complete deglaciation of the main valley so that when Allen Glacier expanded to its present form it had an open valley in which to spread out. This is suggested by the symmetrical mushroom head of this bulb glacier, which is the most perfect illustration of this type of glacier thus far described. The date of this period of expansion is not known except that it was before 1848, as-the vegetation proves. The time since this expansion seems to have been one of such slow movement in the valley glacier that the ice brought forward from the snowfields is practically all removed by vertical ablation in the interior flat. That the glacier is moving forward and is not "dead," as is commonly believed in Alaska, is necessitated by the snow supply in the mountains and proved by the crevassing in the portion of the-interior flat adjacent to the valley glacier. The movement is very slow and does not seem to appreciably affect the outer, or eastern, portion of the bulb at all, for the vegetation has grown in the 4 or 5 feet of ablation moraine soil for more than half a century. When we visited the Allen Glacier in 1909, bare ice was still visible in six of the railway cuts on the glacier terminus. A year later none of these exposures showed ice. The-level railway grade, which had been produced in 1909 by blasting out a shelf in the glacier ice and levelling it up with morainic ballast, was seriously modified in 1910. T?he slumping of the grade, due to melting of the ice beneath, amounted to 2 or 3 feet in many places and, at the maximum, amounted to over 7 feet. Slumping at the water's-edge showed that in 1910 the ice still extended out beneath Copper River. The largest area of slumping beneath the railway grade was 50 by 120 feet. Slumping had commenced here in 1909 and ice was then visible beneath the rails. The settling at this point amounted to 6 feet at one time of observation in 1910, although there had been some-filling in the meantime by the railway section gang. 1 Van Hise, C. R., Principles of North American Pre-Cambrian Geology, 16th Ann. Rept., U. S. GeoL Survey, Part 1,18D6, p. 600.